Abstract
Teleconnections were used to link three El Niño–Southern Oscillation (ENSO) parameters to winter (December–February) cyclone frequencies over the United States during the 1949–96 period. Since each ENSO event is not exactly the same, small subsets of ENSO events were examined in addition to the more common composite ENSO event. Mean winter cyclone frequencies, derived by counting cyclones passing through 30, 5° latitude equal-area circles located in a grid from 70° to 120°W and 30° to 50°N were determined for classes of El Niños and La Niñas based on 1) the intensity of the equatorial Pacific sea surface temperature anomaly, 2) the intensity of the Tahiti–Darwin sea level pressure anomaly, and 3) the location of the 28°C isotherm. The average cyclone count for each class of El Niño and La Niña was compared to the average count for winters when no ENSO event occurred. Expected differences in cyclone frequency patterns when comparing an average of all El Niño winters to all La Niña winters were found; however, large pattern differences were also determined when comparing winters with strong El Niños to moderate–weak El Niños and similarly for La Niñas. Significant differences in number of cyclones were found in 8 of 30 circles located in the Pacific Northwest, the Great Lakes, New England, and the Southeast. The differences found in the cyclone frequency patterns for El Niños and La Niñas of different intensities and locations indicated that using a composite of all El Niños or La Niñas may provide misleading information while examination of each of these parameters independently may assist in the preparation of long-range climate predictions.